Services on Demand
Journal
Article
Indicators
- Cited by SciELO
- Access statistics
Related links
- Similars in SciELO
Share
RISTI - Revista Ibérica de Sistemas e Tecnologias de Informação
Print version ISSN 1646-9895
Abstract
CASTILLO-ROJAS, Wilson; MEDINA-QUISPE, Fernando and VEGA-DAMKE, Juan. New Visualization Scheme for Cluster Models in Data Mining. RISTI [online]. 2017, n.21, pp.67-80. ISSN 1646-9895. https://doi.org/10.17013/risti.21.67-84.
The article proposes the design and implementation of a visualization scheme for cluster models, in the context of a data-mining process. In general, a good cluster model is not difficult to interpret, but its visual representation becomes complex when the data set is of high volume, density and dimensionality. In this type of case, it's necessary to have an appropriate visualization scheme. The visual schema proposed in this work is called VIMC, and is based on four characteristics: interactive visualization, data-mining techniques combination, ad-hoc graphic artifacts, and use of metrics. The considered metrics allow to compare components of different clusters, which in turn helps to understand the composition of the groups. Through the implementation of a visual web environment, and an evaluation of 23 users, positive results are achieved on the utility of this visualization scheme.
Keywords : Cluster Visualization; model visualization for data mining; data mining visualization; interactive visualization of models; data visualization schemes.